setwd("C:/Users/Laura/3D Objects/Thesis")
rm(list=ls())
check.packages <- function(pkg){
new.pkg <- pkg[!(pkg %in% installed.packages()[, "Package"])]
if (length(new.pkg))
install.packages(new.pkg, dependencies = TRUE,
repos = "https://cran.rstudio.com")
sapply(pkg, require, character.only = TRUE)
}
check.packages(c("tidyverse","tidyr", "dplyr","data.table","readxl","haven", "estimatr", "plm", "stargazer", "texreg", "zoo","ncdf4", "chron", "lattice", "RColorBrewer", "lubridate", "padr", "priceR", "tframePlus", "maditr", "gtrendsR","zoo" ,"TSstudio","CausalImpact"))
tidyverse tidyr dplyr data.table readxl haven
TRUE TRUE TRUE TRUE TRUE TRUE
estimatr plm stargazer texreg zoo ncdf4
TRUE TRUE TRUE TRUE TRUE TRUE
chron lattice RColorBrewer lubridate padr priceR
TRUE TRUE TRUE TRUE TRUE TRUE
tframePlus maditr gtrendsR zoo TSstudio CausalImpact
TRUE TRUE TRUE TRUE TRUE TRUE
FOIData<-read_excel("FOIData2017.xlsx", sheet="Table 1")
FOI_NoNationality<-FOIData%>%
dplyr::select(-Nationality, -Year)
FOIAgg_Methods<-FOI_NoNationality%>%
group_by(`Year and Month`, `Method of Entry`)%>%
summarise(Detections=sum(`Number of Detections`))%>%
ungroup()
FOIAgg_Total<-FOI_NoNationality%>%
group_by(`Year and Month`)%>%
summarise(Detections=sum(`Number of Detections`))%>%
ungroup()
FOITotal1<-FOIAgg_Total%>%
separate(`Year and Month`, c("Year", "Month"), sep=" ")%>%
mutate(Year=as.integer(Year), Month=as.integer(Month))
FOITotal <- FOITotal1[-73,]
FOITotal<-FOITotal%>%
mutate(Year=as.character(Year), Month=as.character(Month))
FOITotal$Month <- sprintf("%02d", as.numeric(FOITotal$Month))
FOIMethods1<-FOIAgg_Methods%>%
separate(`Year and Month`, c("Year", "Month"), sep=" ")%>%
mutate(Year=as.integer(Year), Month=as.integer(Month))
FOI_Inad<-FOIMethods1%>%
filter(`Method of Entry`=="Inadequately documented air arrivals")%>%
complete(Year =seq(min(2017), max(2022), 1L) , Month = seq(min(1), max(12), 1L))%>%
replace_na(list(`Method of Entry` = "Inadequately documented air arrivals", Detections=0))
FOI_Ports<-FOIMethods1%>%
filter(`Method of Entry`=="Recorded detections at UK ports")%>%
complete(Year =seq(min(2017), max(2022), 1L) , Month = seq(min(1), max(12), 1L))%>%
replace_na(list(`Method of Entry` = "Recorded detections at UK ports", Detections=0))
FOI_InUK<-FOIMethods1%>%
filter(`Method of Entry`=="Recorded detections in the UK")%>%
complete(Year =seq(min(2017), max(2022), 1L) , Month = seq(min(1), max(12), 1L))%>%
replace_na(list(`Method of Entry` = "Recorded detections in the UK", Detections=0))
FOI_SB<-FOIMethods1%>%
filter(`Method of Entry`=="Small boat arrivals")%>%
complete(Year =seq(min(2017), max(2022), 1L) , Month = seq(min(1), max(12), 1L))%>%
replace_na(list(`Method of Entry` = "Small boat arrivals", Detections=0))
FOIMethods2<-bind_rows(FOI_Inad, FOI_Ports, FOI_InUK, FOI_SB)%>%
mutate(Year=as.character(Year), Month=as.character(Month))
FOIMethods2$Month <- sprintf("%02d", as.numeric(FOIMethods2$Month))
FOIMethods<- FOIMethods2%>%dcast(Year+Month ~ `Method of Entry`)
FOI_Complete<-merge(FOITotal, FOIMethods, by=c("Year", "Month"))
CE1<-read_excel("Conflict_Events1.xlsx")
VaC<-CE1 %>%
filter(event_type == 'Violence against civilians') %>%
group_by(country) %>%
dplyr::select(event_date, event_type, country)%>%
ungroup()%>%
filter(country!='United Arab Emirates')%>%filter(country!='Saudi Arabia')%>%filter(country!='Qatar')%>%filter(country!='Oman')
VaC$Date <- as.Date(VaC$event_date, "%Y/%m/%d")
setDT(VaC)
VaC_counted<-VaC[, .N, by=.(Date, country)]
VaC_counted1<-VaC_counted%>%
rename(VaC_Occurences=N, Country=country) %>%
na.omit()
VaC_padded<-pad(VaC_counted1, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
VaC_padded<-replace(VaC_padded,is.na(VaC_padded), 0)
VaC_padded$Date<-as.character(VaC_padded$Date)
VaC_reshaped<-dcast(VaC_padded, Date ~ Country)
VaC<-VaC_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(VaC_Afg=Afghanistan,VaC_Alb=Albania, VaC_Alg=Algeria,VaC_Bah=Bahrain,VaC_Dji=Djibouti, VaC_Egp=Egypt, VaC_Eri=Eritrea, VaC_Eth=Ethiopia, Vac_Geo=Georgia, VaC_Ind=India, VaC_Iran=Iran, VaC_Iraq=Iraq, VaC_Isr=Israel,VaC_Jor=Jordan,VaC_Leb=Lebanon, VaC_Lib=Libya, VaC_Mor=Morocco, VaC_Pak=Pakistan, VaC_Sri=`Sri Lanka`, VaC_Sud=Sudan, VaC_Syr=Syria, VaC_Tun=Tunisia, VaC_Turk=Turkey, VaC_Viet=Vietnam, VaC_Yem=Yemen)
Protests<-CE1 %>%
filter(event_type == 'Protests') %>%
group_by(country) %>%
dplyr::select(event_date, event_type, country)%>%
filter(country!='United Arab Emirates')%>%filter(country!='Saudi Arabia')%>%filter(country!='Qatar')%>%filter(country!='Oman')
Protests$Date <- as.Date(Protests$event_date, "%Y/%m/%d")
setDT(Protests)
Protests_counted<-Protests[, .N, by=.(Date, country)]
Protests_counted1<-Protests_counted%>%
rename(Protest_Occurences=N, Country=country) %>%
na.omit()
Protests_padded<-pad(Protests_counted1, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
Protests_padded<-replace(Protests_padded,is.na(Protests_padded), 0)
Protests_padded$Date<-as.character(Protests_padded$Date)
Protests_reshaped<-dcast(Protests_padded, Date ~ Country)
Protests<-Protests_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(Pro_Afg=Afghanistan,Pro_Alb=Albania, Pro_Alg=Algeria,Pro_Bah=Bahrain,Pro_Dji=Djibouti, Pro_Egp=Egypt, Pro_Eri=Eritrea, Pro_Eth=Ethiopia, Pro_Geo=Georgia, Pro_Ind=India, Pro_Iran=Iran, Pro_Iraq=Iraq, Pro_Isr=Israel,Pro_Jor=Jordan,Pro_Kuw=Kuwait,Pro_Leb=Lebanon, Pro_Lib=Libya, Pro_Mor=Morocco, Pro_Pak=Pakistan, Pro_Sri=`Sri Lanka`, Pro_Sud=Sudan, Pro_Syr=Syria, Pro_Tun=Tunisia, Pro_Turk=Turkey, Pro_Viet=Vietnam, Pro_Yem=Yemen)
StrDevs<-CE1 %>%
filter(event_type == 'Strategic developments') %>%
group_by(country) %>%
dplyr::select(event_date, event_type, country)%>%
filter(country!='United Arab Emirates')%>%filter(country!='Saudi Arabia')%>%filter(country!='Qatar')%>%filter(country!='Oman')
StrDevs$Date <- as.Date(StrDevs$event_date, "%Y/%m/%d")
setDT(StrDevs)
StrDevs_counted<-StrDevs[, .N, by=.(Date, country)]
StrDevs_counted1<-StrDevs_counted%>%
rename(StrDevs_Occurences=N, Country=country) %>%
na.omit()
StrDevs_padded<-pad(StrDevs_counted1, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
StrDevs_padded<-replace(StrDevs_padded, is.na(StrDevs_padded), 0)
StrDevs_padded$Date<-as.character(StrDevs_padded$Date)
StrDevs_reshaped<-dcast(StrDevs_padded, Date ~ Country)
StrDevs<-StrDevs_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(SD_Afg=Afghanistan,SD_Alb=Albania, SD_Alg=Algeria,SD_Bah=Bahrain,SD_Dji=Djibouti, SD_Egp=Egypt, SD_Eri=Eritrea, SD_Eth=Ethiopia, SD_Geo=Georgia, SD_Ind=India, SD_Iran=Iran, SD_Iraq=Iraq, SD_Isr=Israel,SD_Jor=Jordan,SD_Kuw=Kuwait,SD_Leb=Lebanon, SD_Lib=Libya, SD_Mor=Morocco, SD_Pak=Pakistan, SD_Sri=`Sri Lanka`, SD_Sud=Sudan, SD_Syr=Syria, SD_Tun=Tunisia, SD_Turk=Turkey, SD_Viet=Vietnam, SD_Yem=Yemen)
Explosions<-CE1 %>%
filter(event_type == 'Explosions/Remote violence') %>%
group_by(country) %>%
dplyr::select(event_date, event_type, country)%>%
ungroup()%>%
filter(country!='United Arab Emirates')%>%filter(country!='Saudi Arabia')%>%filter(country!='Qatar')%>%filter(country!='Oman')
Explosions$Date <- as.Date(Explosions$event_date, "%Y/%m/%d")
setDT(Explosions)
Explosions_counted<-Explosions[, .N, by=.(Date, country)]
Explosions_counted1<-Explosions_counted%>%
rename(Explosions_Occurences=N, Country=country) %>%
na.omit()
Explosions_padded<-pad(Explosions_counted1, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
Explosions_padded<-replace(Explosions_padded, is.na(Explosions_padded), 0)
Explosions_padded$Date<-as.character(Explosions_padded$Date)
Explosions_reshaped<-dcast(Explosions_padded, Date ~ Country)
Explosions<-Explosions_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(Expl_Afg=Afghanistan,Expl_Alb=Albania, Expl_Alg=Algeria,Expl_Bah=Bahrain, Expl_Egp=Egypt, Expl_Eri=Eritrea, Expl_Eth=Ethiopia, Expl_Geo=Georgia, Expl_Ind=India, Expl_Iran=Iran, Expl_Iraq=Iraq, Expl_Isr=Israel,Expl_Jor=Jordan,Expl_Leb=Lebanon, Expl_Lib=Libya, Expl_Mor=Morocco, Expl_Pak=Pakistan, Expl_Sri=`Sri Lanka`, Expl_Sud=Sudan, Expl_Syr=Syria, Expl_Tun=Tunisia, Expl_Turk=Turkey, Expl_Viet=Vietnam, Expl_Yem=Yemen)
Battles<-CE1 %>%
filter(event_type == 'Battles') %>%
group_by(country) %>%
dplyr::select(event_date, event_type, country)%>%
ungroup()%>%
filter(country!='United Arab Emirates')%>%filter(country!='Saudi Arabia')%>%filter(country!='Qatar')
Battles$Date <- as.Date(Battles$event_date, "%Y/%m/%d")
setDT(Battles)
Battles_counted<-Battles[, .N, by=.(Date, country)]
Battles_counted1<-Battles_counted%>%
rename(Battles_Occurences=N, Country=country) %>%
na.omit()
Battles_padded<-pad(Battles_counted1, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
Battles_padded<-replace(Battles_padded, is.na(Battles_padded), 0)
Battles_padded$Date<-as.character(Battles_padded$Date)
Battles_reshaped<-dcast(Battles_padded, Date ~ Country)
Battles<-Battles_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(Batt_Afg=Afghanistan,Batt_Alb=Albania, Batt_Alg=Algeria,Batt_Bah=Bahrain,Batt_Dji=Djibouti, Batt_Egp=Egypt, Batt_Eri=Eritrea, Batt_Eth=Ethiopia, Batt_Geo=Georgia, Batt_Ind=India, Batt_Iran=Iran, Batt_Iraq=Iraq, Batt_Isr=Israel,Batt_Jor=Jordan,Batt_Leb=Lebanon, Batt_Lib=Libya, Batt_Mor=Morocco, Batt_Pak=Pakistan, Batt_Sri=`Sri Lanka`, Batt_Sud=Sudan, Batt_Syr=Syria, Batt_Tun=Tunisia, Batt_Turk=Turkey, Batt_Viet=Vietnam, Batt_Yem=Yemen)
Riots<-CE1 %>%
filter(event_type == 'Riots') %>%
group_by(country) %>%
dplyr::select(event_date, event_type, country)%>%
filter(country!='United Arab Emirates')%>%filter(country!='Saudi Arabia')%>%filter(country!='Qatar')%>%filter(country!='Oman')
Riots$Date <- as.Date(Riots$event_date, "%Y/%m/%d")
setDT(Riots)
Riots_counted<-Riots[, .N, by=.(Date, country)]
Riots_counted1<-Riots_counted%>%
rename(Riots_Occurences=N, Country=country) %>%
na.omit()
Riots_padded<-pad(Riots_counted1, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
Riots_padded<-replace(Riots_padded, is.na(Riots_padded), 0)
Riots_padded$Date<-as.character(Riots_padded$Date)
Riots_reshaped<-dcast(Riots_padded, Date ~ Country)
Riots<-Riots_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(Riot_Afg=Afghanistan,Riot_Alb=Albania, Riot_Alg=Algeria,Riot_Bah=Bahrain,Riot_Dji=Djibouti, Riot_Egp=Egypt, Riot_Eri=Eritrea, Riot_Eth=Ethiopia, Riot_Geo=Georgia, Riot_Ind=India, Riot_Iran=Iran, Riot_Iraq=Iraq, Riot_Isr=Israel,Riot_Jor=Jordan,Riot_Kuw=Kuwait,Riot_Leb=Lebanon, Riot_Lib=Libya, Riot_Mor=Morocco, Riot_Pak=Pakistan, Riot_Sri=`Sri Lanka`, Riot_Sud=Sudan, Riot_Syr=Syria, Riot_Tun=Tunisia, Riot_Turk=Turkey, Riot_Viet=Vietnam, Riot_Yem=Yemen)
Merged<-merge(VaC, Protests, "Date")
Merged<-merge(Merged, StrDevs, by="Date")
Merged<-merge(Merged, Explosions, by="Date")
Merged<-merge(Merged, Battles, by="Date")
Conflict_Data<-merge(Merged, Riots, by="Date")
Conflict_Monthly <- Conflict_Data # Duplicate data
Conflict_Monthly$Year <- strftime(Conflict_Monthly$Date, "%Y") # Create year column
Conflict_Monthly$Month <- strftime(Conflict_Monthly$Date, "%m") # Create month column
Conflict_Agg <- aggregate(. ~ Month + Year,
Conflict_Monthly,
FUN = sum)
Conflict_Agg<-Conflict_Agg%>%dplyr::select(!Date)
Disasters_Raw <- read_excel("Disasters1.xlsx")
Disasters_selected<- Disasters_Raw %>%
dplyr::select("Start Year", "Start Month", "Start Day","Country","Disaster Group") %>%
filter(Country!='United Arab Emirates (the)')%>%filter(Country!='Saudi Arabia')%>%filter(Country!='Qatar')%>%filter(Country!='Oman')
Disasters_selected <- replace(Disasters_selected, is.na(Disasters_selected), 1)#giving this event an arbitrary start day so the date function works
Disasters_selectedDate<-Disasters_selected %>% mutate(Date = make_date(year = `Start Year`, month = `Start Month`, day = `Start Day`))
###Split: Total Disasters####
setDT(Disasters_selectedDate)
Disasters_counted<-Disasters_selectedDate[, .N, by=.(Date, Country)] %>%
rename(DisasterOccurence=N)
Disasters_padded<-pad(Disasters_counted, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
Disasters_padded1 <- Disasters_padded %>% replace(is.na(.), 0)
Disasters_padded1$Date<-as.character(Disasters_padded1$Date)
Disasters_reshaped<-dcast(Disasters_padded1, Date ~ Country)
Disasters<-Disasters_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(Dis_Afg=Afghanistan,Dis_Alb=Albania, Dis_Alg=Algeria,Dis_Dji=Djibouti, Dis_Egp=Egypt, Dis_Eri=Eritrea, Dis_Eth=Ethiopia, Dis_Geo=Georgia, Dis_Ind=India, Dis_Iran=`Iran (Islamic Republic of)`, Dis_Iraq=Iraq, Dis_Isr=Israel,Dis_Jor=Jordan,Dis_Kuw=Kuwait,Dis_Leb=Lebanon, Dis_Lib=Libya, Dis_Mor=Morocco, Dis_Pak=Pakistan, Dis_Sri=`Sri Lanka`, Dis_Sud=`Sudan (the)`, Dis_Syr=`Syrian Arab Republic`, Dis_Tun=Tunisia, Dis_Turk=Turkey, Dis_Viet=`Viet Nam`, Dis_Yem=Yemen, Dis_Pal=`Palestine, State of`, Dis_SoSud=`South Sudan`)
Natural<-Disasters_selectedDate %>%
filter(`Disaster Group` == 'Natural') %>%
dplyr::select(Date, Country, `Disaster Group`)
setDT(Natural)
Natural_counted<-Natural[, .N, by=.(Date, Country)] %>%
rename(NaturalOccurence=N)
Natural_padded<-pad(Natural_counted, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
Natural_padded <- replace(Natural_padded, is.na(Natural_padded), 0)
Natural_padded$Date<-as.character(Natural_padded$Date)
Natural_reshaped<-dcast(Natural_padded, Date ~ Country)
Natural<-Natural_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(Nat_Afg=Afghanistan,Nat_Alb=Albania, Nat_Alg=Algeria,Nat_Dji=Djibouti, Nat_Egp=Egypt, Nat_Eri=Eritrea, Nat_Eth=Ethiopia, Nat_Geo=Georgia, Nat_Ind=India, Nat_Iran=`Iran (Islamic Republic of)`, Nat_Iraq=Iraq, Nat_Isr=Israel,Nat_Jor=Jordan,Nat_Kuw=Kuwait,Nat_Leb=Lebanon, Nat_Lib=Libya, Nat_Mor=Morocco, Nat_Pak=Pakistan, Nat_Sri=`Sri Lanka`, Nat_Sud=`Sudan (the)`, Nat_Syr=`Syrian Arab Republic`, Nat_Tun=Tunisia, Nat_Turk=Turkey, Nat_Viet=`Viet Nam`, Nat_Yem=Yemen, Nat_Pal=`Palestine, State of`, Nat_SoSud=`South Sudan`)
Technological<-Disasters_selectedDate %>%
filter(`Disaster Group` == 'Technological') %>%
dplyr::select(Date, Country, `Disaster Group`)
setDT(Technological)
Technological_counted<-Technological[, .N, by=.(Date,Country)] %>%
rename(TechnologicalOccurences=N)
Tech_padded<-pad(Technological_counted, group="Country", start_val =as.Date("2017-01-01" ), end_val = as.Date("2022-12-31"))
Tech_padded <- replace(Tech_padded, is.na(Tech_padded), 0)
Tech_padded$Date<-as.character(Tech_padded$Date)
Tech_reshaped<-dcast(Tech_padded, Date ~ Country)
Technological<-Tech_reshaped%>%
mutate(Date=as.Date(Date))%>%
rename(Tech_Afg=Afghanistan, Tech_Alg=Algeria,Tech_Dji=Djibouti, Tech_Egp=Egypt, Tech_Eth=Ethiopia, Tech_Geo=Georgia, Tech_Ind=India, Tech_Iran=`Iran (Islamic Republic of)`, Tech_Iraq=Iraq, Tech_Isr=Israel,Tech_Jor=Jordan,Tech_Kuw=Kuwait,Tech_Leb=Lebanon, Tech_Lib=Libya, Tech_Mor=Morocco, Tech_Pak=Pakistan , Tech_Sri=`Sri Lanka`, Tech_Sud=`Sudan (the)`, Tech_Syr=`Syrian Arab Republic`, Tech_Tun=Tunisia, Tech_Turk=Turkey, Tech_Viet=`Viet Nam`, Tech_Yem=Yemen, Tech_Pal=`Palestine, State of`, Tech_SoSud=`South Sudan`)
#Merging Disasters####
Disasters_Merged<-merge(Natural, Technological, by="Date" )
Disasters_Merged<-merge(Disasters_Merged, Disasters, by="Date" )
####Monthly Aggregation
Disasters_Monthly <- Disasters_Merged
Disasters_Monthly$Year <- strftime(Disasters_Monthly$Date, "%Y")
Disasters_Monthly$Month <- strftime(Disasters_Monthly$Date, "%m")
Disasters_Agg <- aggregate(. ~ Month + Year,
Disasters_Monthly,
FUN = sum)
Disasters_Agg<-Disasters_Agg%>%
dplyr::select(!Date)
#Monthly, no country division
Commodities_Raw <- read_excel("Commodity_Prices.xls") %>%
filter(Commodity=="2017M1"|Commodity=="2017M2"|Commodity=="2017M3"|Commodity=="2017M4"|Commodity=="2017M5"|Commodity=="2017M6"|Commodity=="2017M7"|Commodity=="2017M8"|Commodity=="2017M9"|Commodity=="2017M10"|Commodity=="2017M11"|Commodity=="2017M12"|
Commodity=="2018M1"|Commodity=="2018M2"|Commodity=="2018M3"|Commodity=="2018M4"|Commodity=="2018M5"|Commodity=="2018M6"|Commodity=="2018M7"|Commodity=="2018M8"|Commodity=="2018M9"|Commodity=="2018M10"|Commodity=="2018M11"|Commodity=="2018M12"|
Commodity=="2019M1"|Commodity=="2019M2"|Commodity=="2019M3"|Commodity=="2019M4"|Commodity=="2019M5"|Commodity=="2019M6"|Commodity=="2019M7"|Commodity=="2019M8"|Commodity=="2019M9"|Commodity=="2019M10"|Commodity=="2019M11"|Commodity=="2019M12"|
Commodity=="2020M1"|Commodity=="2020M2"|Commodity=="2020M3"|Commodity=="2020M4"|Commodity=="2020M5"|Commodity=="2020M6"|Commodity=="2020M7"|Commodity=="2020M8"|Commodity=="2020M9"|Commodity=="2020M10"|Commodity=="2020M11"|Commodity=="2020M12"|
Commodity=="2021M1"|Commodity=="2021M2"|Commodity=="2021M3"|Commodity=="2021M4"|Commodity=="2021M5"|Commodity=="2021M6"|Commodity=="2021M7"|Commodity=="2021M8"|Commodity=="2021M9"|Commodity=="2021M10"|Commodity=="2021M11"|Commodity=="2021M12"|
Commodity=="2022M1"|Commodity=="2022M2"|Commodity=="2022M3"|Commodity=="2022M4"|Commodity=="2022M5"|Commodity=="2022M6"|Commodity=="2022M7"|Commodity=="2022M8"|Commodity=="2022M9"|Commodity=="2022M10"|Commodity=="2022M11"|Commodity=="2022M12")%>%
dplyr::select('Commodity','PALLFNF', 'PNFUEL', 'PNRG', 'PAGRI', 'PMETA', 'PFERT' )
Commodities_Dated<-Commodities_Raw %>%
separate(Commodity, c("Year", "Month"), "M")
Commodities_Dated$Month <- sprintf("%02d", as.numeric(Commodities_Dated$Month))
UK_Unemployment <- read_excel("UK_Unemployment_EndingDec2022.xls") %>%
separate(Title, c("Year", "Month"), " ")
UK_Unemployment[UK_Unemployment == "JAN"] <- "1"
UK_Unemployment[UK_Unemployment == "FEB"] <- "2"
UK_Unemployment[UK_Unemployment == "MAR"] <- "3"
UK_Unemployment[UK_Unemployment == "APR"] <- "4"
UK_Unemployment[UK_Unemployment == "MAY"] <- "5"
UK_Unemployment[UK_Unemployment == "JUN"] <- "6"
UK_Unemployment[UK_Unemployment == "JUL"] <- "7"
UK_Unemployment[UK_Unemployment == "AUG"] <- "8"
UK_Unemployment[UK_Unemployment == "SEP"] <- "9"
UK_Unemployment[UK_Unemployment == "OCT"] <- "10"
UK_Unemployment[UK_Unemployment == "NOV"] <- "11"
UK_Unemployment[UK_Unemployment == "DEC"] <- "12"
UK_Unemployment$Year <- as.integer(UK_Unemployment$Year)
UK_Unemployment$Month <- as.integer(UK_Unemployment$Month)
UK_Unemployment$Date <- as.yearmon(paste(UK_Unemployment$Year, UK_Unemployment$Month), "%Y %m")
UK_Unemployment<-UK_Unemployment%>%
mutate(Year=as.character(Year), Month=as.character(Month))%>%
dplyr::select(-Date)
UK_Unemployment$Month <- sprintf("%02d", as.numeric(UK_Unemployment$Month))
Channel_Weather <- read_excel("Channel_Weather.xlsx")
Channel_Weather$datetime <- as.Date(Channel_Weather$datetime)
Channel_Weather_Monthly<-Channel_Weather
Channel_Weather_Monthly$Date <- floor_date(Channel_Weather_Monthly$datetime, "month")
Channel_Weather_Monthly1<- Channel_Weather_Monthly%>%dplyr::select(!datetime)
Channel_Weather_Monthly2<-Channel_Weather_Monthly1 %>%
group_by(Date) %>%
summarize(mean_temp = mean(temp), mean_min = mean(tempmin), mean_max = mean(tempmax), mean_precip = mean(precip), mean_windgust = mean(windgust), mean_wspeed = mean(windspeed), mean_winddir = mean(winddir), mean_visibility = mean(visibility)) %>%
mutate(Year = strftime(Date, "%Y"), Month=strftime(Date, "%m"))%>%ungroup()
Channel_Weather_Monthly3<-Channel_Weather_Monthly2%>% dplyr::select(Year, Month, mean_temp, mean_precip, mean_wspeed)
Freedom_raw<-read_excel("Freedom_PR-CR.xlsx")
Freedom<- Freedom_raw%>%
dplyr::select("Country/Territory", "Edition", "PR rating", "CL rating", "Total")%>%
rename(Country='Country/Territory', Year=Edition)%>%
filter(Country=='Afghanistan'|Country=='Albania'|Country=='Iraq'|Country=='Iran'|Country=='Syria'|Country=='Eritrea'|Country=='Sudan'|Country=='Egypt'|Country=='Turkey'|Country=='Ethiopia'|Country=='Vietnam'|Country=='Kuwait'|Country=='India'|Country=='Georgia'|Country=='Pakistan'|Country=='Sri Lanka'|Country=='Yemen'|Country=='Libya'|Country=='Algeria'|Country=='Djibouti'|Country=='Bahrain'|Country=='Israel'|Country=='Jordan'|Country=='Lebanon'|Country=='Morocco'|Country=='Tunisia'|Country=='Yemen'|Country=='Palestine')%>%
mutate(Year=as.integer(Year))%>%
filter(Year=='2017'|Year=='2018'|Year=='2019'|Year=='2020'|Year=='2021'|Year=='2022')
Freedom_rep <- Freedom %>% slice(rep(1:n(), each = 12)) %>%
mutate(Month=rep(c(1,2,3,4,5, 6, 7, 8, 9, 10, 11, 12),times=156))%>%
mutate(Year=as.character(Year), Month=as.character(Month))
Freedom_rep$Month <- sprintf("%02d", as.numeric(Freedom_rep$Month))
PR1<-Freedom_rep%>%
dplyr::select(Year,Month, Country, `PR rating`)
PR_reshaped<-dcast(PR1, Year+Month ~ Country)
PR2<-PR_reshaped%>%
rename(PR_Afg=Afghanistan,PR_Alb=Albania, PR_Alg=Algeria,PR_Dji=Djibouti, PR_Egp=Egypt, PR_Eri=Eritrea, PR_Eth=Ethiopia, PR_Geo=Georgia, PR_Ind=India, PR_Iran=`Iran`, PR_Iraq=Iraq, PR_Isr=Israel,PR_Jor=Jordan,PR_Kuw=Kuwait,PR_Leb=Lebanon, PR_Lib=Libya, PR_Mor=Morocco, PR_Pak=Pakistan, PR_Sri=`Sri Lanka`, PR_Sud=`Sudan`, PR_Syr=`Syria`, PR_Tun=Tunisia, PR_Turk=Turkey, PR_Viet=`Vietnam`, PR_Yem=Yemen, PR_Bah=Bahrain)
CL1<-Freedom_rep%>%
dplyr::select(Year,Month, Country, `CL rating`)
CL_reshaped<-dcast(CL1, Year+Month ~ Country)
CL2<-CL_reshaped%>%
rename(CL_Afg=Afghanistan,CL_Alb=Albania, CL_Alg=Algeria,CL_Dji=Djibouti, CL_Egp=Egypt, CL_Eri=Eritrea, CL_Eth=Ethiopia, CL_Geo=Georgia, CL_Ind=India, CL_Iran=`Iran`, CL_Iraq=Iraq, CL_Isr=Israel,CL_Jor=Jordan,CL_Kuw=Kuwait,CL_Leb=Lebanon, CL_Lib=Libya, CL_Mor=Morocco, CL_Pak=Pakistan, CL_Sri=`Sri Lanka`, CL_Sud=`Sudan`, CL_Syr=`Syria`, CL_Tun=Tunisia, CL_Turk=Turkey, CL_Viet=`Vietnam`, CL_Yem=Yemen, CL_Bah=Bahrain)
PRCL<-merge(PR2, CL2, by=c("Year", "Month"))
Afghanistan<-historical_exchange_rates("GBP", "AFN", "2017-01-01", "2022-12-31")
Albania<-historical_exchange_rates("GBP", "ALL", "2017-01-01", "2022-12-31")
Algeria<-historical_exchange_rates("GBP", "DZD", "2017-01-01", "2022-12-31")
Egypt<-historical_exchange_rates("GBP", "EGP", "2017-01-01", "2022-12-31")
Eritrea<-historical_exchange_rates("GBP", "ERN", "2017-01-01", "2022-12-31")
Ethiopia<-historical_exchange_rates("GBP", "ETB", "2017-01-01", "2022-12-31")
Georgia<-historical_exchange_rates("GBP", "GEL", "2017-01-01", "2022-12-31")
India<-historical_exchange_rates("GBP", "INR", "2017-01-01", "2022-12-31")
Iran<-historical_exchange_rates("GBP", "IRR", "2017-01-01", "2022-12-31")
Iraq<-historical_exchange_rates("GBP", "IQD", "2017-01-01", "2022-12-31")
Kuwait<-historical_exchange_rates("GBP", "KWD", "2017-01-01", "2022-12-31")
Libya<-historical_exchange_rates("GBP", "LYD", "2017-01-01", "2022-12-31")
Pakistan<-historical_exchange_rates("GBP", "PKR", "2017-01-01", "2022-12-31")
SriLanka<-historical_exchange_rates("GBP", "LKR", "2017-01-01", "2022-12-31")
Sudan<-historical_exchange_rates("GBP", "SDG", "2017-01-01", "2022-12-31")
Syria<-historical_exchange_rates("GBP", "SYP", "2017-01-01", "2022-12-31")
Turkey<-historical_exchange_rates("GBP", "TRY", "2017-01-01", "2022-12-31")
Vietnam<-historical_exchange_rates("GBP", "VND", "2017-01-01", "2022-12-31")
Yemen<-historical_exchange_rates("GBP", "YER", "2017-01-01", "2022-12-31")
Euro<-historical_exchange_rates("GBP", "EUR", "2017-01-01", "2022-12-31")
Bahrain<-historical_exchange_rates("GBP", "BHD", "2017-01-01", "2022-12-31")
Djibouti<-historical_exchange_rates("GBP", "DJF", "2017-01-01", "2022-12-31")
Israel<-historical_exchange_rates("GBP", "ILS", "2017-01-01", "2022-12-31")
Jordan<-historical_exchange_rates("GBP", "JOD", "2017-01-01", "2022-12-31")
Lebanon<-historical_exchange_rates("GBP", "LBP", "2017-01-01", "2022-12-31")
Morocco<-historical_exchange_rates("GBP", "MAD", "2017-01-01", "2022-12-31")
Tunisia<-historical_exchange_rates("GBP", "TND", "2017-01-01", "2022-12-31")
merge<-merge(Afghanistan, Albania, by="date")
merge<-merge(merge, Algeria, by ="date")
merge<-merge(merge, Egypt, by ="date")
merge<-merge(merge, Eritrea, by ="date")
merge<-merge(merge, Ethiopia, by ="date")
merge<-merge(merge, Georgia, by ="date")
merge<-merge(merge, India, by ="date")
merge<-merge(merge, Iran, by ="date")
merge<-merge(merge, Iraq, by ="date")
merge<-merge(merge, Kuwait, by ="date")
merge<-merge(merge, Libya, by ="date")
merge<-merge(merge, Pakistan, by ="date")
merge<-merge(merge, SriLanka, by ="date")
merge<-merge(merge, Sudan, by ="date")
merge<-merge(merge, Syria, by ="date")
merge<-merge(merge, Turkey, by ="date")
merge<-merge(merge, Vietnam, by ="date")
merge<-merge(merge, Yemen, by ="date")
merge<-merge(merge, Euro, by ="date")
merge<-merge(merge, Bahrain, by ="date")
merge<-merge(merge, Djibouti, by ="date")
merge<-merge(merge, Israel, by ="date")
merge<-merge(merge, Jordan, by ="date")
merge<-merge(merge, Lebanon, by ="date")
merge<-merge(merge, Morocco, by ="date")
merge<-merge(merge, Tunisia, by ="date")
Exchange_Rates<-merge
Exchange_Rates$Year <- strftime(Exchange_Rates$date, "%Y")
Exchange_Rates$Month <- strftime(Exchange_Rates$date, "%m")
Exchange_Rates_Monthly <- aggregate(cbind(one_GBP_equivalent_to_x_AFN, one_GBP_equivalent_to_x_ALL, one_GBP_equivalent_to_x_DZD, one_GBP_equivalent_to_x_EGP, one_GBP_equivalent_to_x_ERN, one_GBP_equivalent_to_x_ETB, one_GBP_equivalent_to_x_GEL, one_GBP_equivalent_to_x_INR, one_GBP_equivalent_to_x_IRR,one_GBP_equivalent_to_x_IQD, one_GBP_equivalent_to_x_KWD, one_GBP_equivalent_to_x_LYD, one_GBP_equivalent_to_x_PKR, one_GBP_equivalent_to_x_LKR, one_GBP_equivalent_to_x_SDG, one_GBP_equivalent_to_x_SYP, one_GBP_equivalent_to_x_TRY, one_GBP_equivalent_to_x_VND, one_GBP_equivalent_to_x_YER, one_GBP_equivalent_to_x_EUR, one_GBP_equivalent_to_x_BHD, one_GBP_equivalent_to_x_DJF, one_GBP_equivalent_to_x_ILS, one_GBP_equivalent_to_x_JOD, one_GBP_equivalent_to_x_LBP, one_GBP_equivalent_to_x_MAD, one_GBP_equivalent_to_x_TND) ~ Month + Year,
Exchange_Rates,
FUN = mean)
GDPPC_raw<-read_excel("GDPPC.xls")
GDP_selected<-GDPPC_raw%>%
dplyr::select("Country Name","2017", "2018", "2019","2020", "2021")%>%
filter(`Country Name`=="Albania"|`Country Name`=="Afghanistan"|`Country Name`=="Iraq"|`Country Name`=="Iran, Islamic Rep."|`Country Name`=="Syrian Arab Republic"|`Country Name`=="United Kingdom")%>%
rename(Country=`Country Name`)
GDP_melted<-GDP_selected%>%
melt(id = "Country")
GDP_melted1<-GDP_melted%>%
rename(Year=variable, GDPpc=value)
Compared<-GDP_melted1%>%
mutate(UKValue=ifelse(Year=="2017",40621.3345, ifelse(Year=="2018",43306.3083, ifelse(Year=="2019",42747.0805, ifelse(Year=="2020",40318.5576, ifelse(Year=="2021",46510.2828, 0))) ) ))
Compared1<-Compared%>%
mutate(GDPComparison=GDPpc/UKValue)
GDPpc_Final<-Compared1%>%
mutate(Country = if_else(Country == 'Syrian Arab Republic', 'Syria', Country))%>%
mutate(Country = if_else(Country == 'Iran, Islamic Rep.', 'Iran', Country))
covid_trends= gtrends(keyword="covid", gprop ="web",time= "2017-01-01 2022-12-31")
covid_time_trend=covid_trends$interest_over_time
covid_time_trend$Year <- strftime(covid_time_trend$date, "%Y")
covid_time_trend$Month <- strftime(covid_time_trend$date, "%m")
covid_time_trend<-covid_time_trend%>%
mutate(hits=as.numeric(hits))
covid_time_trend<-covid_time_trend%>%
replace(is.na(.),0)
covid_time_trend<-covid_time_trend%>%
dplyr::select(Year, Month, keyword, hits)
covid_aggr1 <- aggregate(hits ~ Month + Year+keyword, covid_time_trend, FUN = mean)
covid_aggr1<-dcast(covid_aggr1, Year+Month~keyword)%>%rename(g_covid=covid)
covide<-read.csv("https://raw.githubusercontent.com/OxCGRT/covid-policy-tracker/master/data/United%20Kingdom/OxCGRT_GBR_differentiated_withnotes_2020.csv")
covid2020<-covide%>%
dplyr::select(Date, C8EV_International.travel.controls)
covide2021<-read.csv("https://raw.githubusercontent.com/OxCGRT/covid-policy-tracker/master/data/United%20Kingdom/OxCGRT_GBR_differentiated_withnotes_2021.csv")
covid2021<-covide2021%>%
dplyr::select(Date, C8EV_International.travel.controls)
covide2022<-read.csv("https://raw.githubusercontent.com/OxCGRT/covid-policy-tracker/master/data/United%20Kingdom/OxCGRT_GBR_differentiated_withnotes_2022.csv")
covid2022<-covide2022%>%
dplyr::select(Date, C8EV_International.travel.controls)
mergecovid<-rbind(covid2020, covid2021)
mergecovid1<-rbind(mergecovid, covid2022)
correcteddate<-mergecovid1%>%
mutate(Date=as.character(Date))%>%
separate(Date, c("Year", "Month", "Day"), sep=cumsum(c(4,2,2)))
aggregatedcovid<-aggregate(C8EV_International.travel.controls ~ Month+Year, correcteddate,FUN = mean)
covid_Filled<-aggregatedcovid%>%
mutate(Year=as.integer(Year), Month=as.integer(Month))%>%
complete(Year =seq(min(2017), max(2022), 1L) , Month = seq(min(1), max(12), 1L))
covid_Final<-replace(covid_Filled, is.na(covid_Filled), 0)
covid_Final$Month <- sprintf("%02d", as.numeric(covid_Final$Month))
FrontexRaw<-read_excel("FrontexArrivals.xlsx")
Afg_Frontex<-FrontexRaw%>%
dplyr::select(-Route, -`Border type or inland`)
melt<-Afg_Frontex%>%
melt(id="Nationality")%>%
dplyr::select(-Nationality)
nextt<-melt%>%
group_by(variable)%>%
summarise(value=sum(value))%>%
ungroup()
Separated<-nextt%>%
mutate(variable=as.character(variable))%>%
separate(variable, c("Month", "Year"), sep=cumsum(c(3,4)))%>%
filter(Year==2017|Year==2018|Year==2019|Year==2020|Year==2021|Year==2022)
Separated[Separated == "JAN"] <- "01"
Separated[Separated == "FEB"] <- "02"
Separated[Separated == "MAR"] <- "03"
Separated[Separated == "APR"] <- "04"
Separated[Separated == "MAY"] <- "05"
Separated[Separated == "JUN"] <- "06"
Separated[Separated == "JUL"] <- "07"
Separated[Separated == "AUG"] <- "08"
Separated[Separated == "SEP"] <- "09"
Separated[Separated == "OCT"] <- "10"
Separated[Separated == "NOV"] <- "11"
Separated[Separated == "DEC"] <- "12"
Frontex<-Separated%>%
rename(FrontexDetections=value)
visa_AF= gtrends(keyword="visa",geo="AF", gprop ="web",time= "2017-01-01 2022-12-31")
visa_AF_time_trend=visa_AF$interest_over_time
visa_AF_time_trend$Year <- strftime(visa_AF_time_trend$date, "%Y")
visa_AF_time_trend$Month <- strftime(visa_AF_time_trend$date, "%m")
visa_AF_time_trend<-visa_AF_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
visa_AF_aggr1 <- aggregate(hits ~ Month + Year+keyword, visa_AF_time_trend, FUN = mean)
visa_AF_aggr1<-dcast(visa_AF_aggr1, Year+Month~keyword)%>%rename(visa_AF=visa)
visa_AL= gtrends(keyword="visa",geo="AL", gprop ="web",time= "2017-01-01 2022-12-31")
visa_AL_time_trend=visa_AL$interest_over_time
visa_AL_time_trend$Year <- strftime(visa_AL_time_trend$date, "%Y")
visa_AL_time_trend$Month <- strftime(visa_AL_time_trend$date, "%m")
visa_AL_time_trend<-visa_AL_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
visa_AL_aggr1 <- aggregate(hits ~ Month + Year+keyword, visa_AL_time_trend, FUN = mean)
visa_AL_aggr1<-dcast(visa_AL_aggr1, Year+Month~keyword)%>%rename(visa_AL=visa)
visa_IR= gtrends(keyword="visa",geo="IR", gprop ="web",time= "2017-01-01 2022-12-31")
visa_IR_time_trend=visa_IR$interest_over_time
visa_IR_time_trend$Year <- strftime(visa_IR_time_trend$date, "%Y")
visa_IR_time_trend$Month <- strftime(visa_IR_time_trend$date, "%m")
visa_IR_time_trend<-visa_IR_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
visa_IR_aggr1 <- aggregate(hits ~ Month + Year+keyword, visa_IR_time_trend, FUN = mean)
visa_IR_aggr1<-dcast(visa_IR_aggr1, Year+Month~keyword)%>%rename(visa_IR=visa)
visa_SY= gtrends(keyword="visa",geo="SY", gprop ="web",time= "2017-01-01 2022-12-31")
visa_SY_time_trend=visa_SY$interest_over_time
visa_SY_time_trend$Year <- strftime(visa_SY_time_trend$date, "%Y")
visa_SY_time_trend$Month <- strftime(visa_SY_time_trend$date, "%m")
visa_SY_time_trend<-visa_SY_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
visa_SY_aggr1 <- aggregate(hits ~ Month + Year+keyword, visa_SY_time_trend, FUN = mean)
visa_SY_aggr1<-dcast(visa_SY_aggr1, Year+Month~keyword)%>%rename(visa_SY=visa)
visa_IQ= gtrends(keyword="visa",geo="IQ", gprop ="web",time= "2017-01-01 2022-12-31")
visa_IQ_time_trend=visa_IQ$interest_over_time
visa_IQ_time_trend$Year <- strftime(visa_IQ_time_trend$date, "%Y")
visa_IQ_time_trend$Month <- strftime(visa_IQ_time_trend$date, "%m")
visa_IQ_time_trend<-visa_IQ_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
visa_IQ_aggr1 <- aggregate(hits ~ Month + Year+keyword, visa_IQ_time_trend, FUN = mean)
visa_IQ_aggr1<-dcast(visa_IQ_aggr1, Year+Month~keyword)%>%rename(visa_IQ=visa)
visamerge<-merge(visa_AF_aggr1, visa_AL_aggr1, by=c("Year", "Month"))
visamerge<-merge(visamerge, visa_IR_aggr1, by=c("Year", "Month"))
visamerge<-merge(visamerge, visa_IQ_aggr1, by=c("Year", "Month"))
visamerge<-merge(visamerge, visa_SY_aggr1, by=c("Year", "Month"))
job_AF= gtrends(keyword="job",geo="AF", gprop ="web",time= "2017-01-01 2022-12-31")
job_AF_time_trend=job_AF$interest_over_time
job_AF_time_trend$Year <- strftime(job_AF_time_trend$date, "%Y")
job_AF_time_trend$Month <- strftime(job_AF_time_trend$date, "%m")
job_AF_time_trend<-job_AF_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
job_AF_aggr1 <- aggregate(hits ~ Month + Year+keyword, job_AF_time_trend, FUN = mean)
job_AF_aggr1<-dcast(job_AF_aggr1, Year+Month~keyword)%>%rename(job_AF=job)
job_AL= gtrends(keyword="job",geo="AL", gprop ="web",time= "2017-01-01 2022-12-31")
job_AL_time_trend=job_AL$interest_over_time
job_AL_time_trend$Year <- strftime(job_AL_time_trend$date, "%Y")
job_AL_time_trend$Month <- strftime(job_AL_time_trend$date, "%m")
job_AL_time_trend<-job_AL_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
job_AL_aggr1 <- aggregate(hits ~ Month + Year+keyword, job_AL_time_trend, FUN = mean)
job_AL_aggr1<-dcast(job_AL_aggr1, Year+Month~keyword)%>%rename(job_AL=job)
job_IR= gtrends(keyword="job",geo="IR", gprop ="web",time= "2017-01-01 2022-12-31")
job_IR_time_trend=job_IR$interest_over_time
job_IR_time_trend$Year <- strftime(job_IR_time_trend$date, "%Y")
job_IR_time_trend$Month <- strftime(job_IR_time_trend$date, "%m")
job_IR_time_trend<-job_IR_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
job_IR_aggr1 <- aggregate(hits ~ Month + Year+keyword, job_IR_time_trend, FUN = mean)
job_IR_aggr1<-dcast(job_IR_aggr1, Year+Month~keyword)%>%rename(job_IR=job)
job_SY= gtrends(keyword="job",geo="SY", gprop ="web",time= "2017-01-01 2022-12-31")
job_SY_time_trend=job_SY$interest_over_time
job_SY_time_trend$Year <- strftime(job_SY_time_trend$date, "%Y")
job_SY_time_trend$Month <- strftime(job_SY_time_trend$date, "%m")
job_SY_time_trend<-job_SY_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
job_SY_aggr1 <- aggregate(hits ~ Month + Year+keyword, job_SY_time_trend, FUN = mean)
job_SY_aggr1<-dcast(job_SY_aggr1, Year+Month~keyword)%>%rename(job_SY=job)
job_IQ= gtrends(keyword="job",geo="IQ", gprop ="web",time= "2017-01-01 2022-12-31")
job_IQ_time_trend=job_IQ$interest_over_time
job_IQ_time_trend$Year <- strftime(job_IQ_time_trend$date, "%Y")
job_IQ_time_trend$Month <- strftime(job_IQ_time_trend$date, "%m")
job_IQ_time_trend<-job_IQ_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
job_IQ_aggr1 <- aggregate(hits ~ Month + Year+keyword, job_IQ_time_trend, FUN = mean)
job_IQ_aggr1<-dcast(job_IQ_aggr1, Year+Month~keyword)%>%rename(job_IQ=job)
jobmerge<-merge(job_AF_aggr1, job_AL_aggr1, by=c("Year", "Month"))
jobmerge<-merge(jobmerge, job_IR_aggr1, by=c("Year", "Month"))
jobmerge<-merge(jobmerge, job_IQ_aggr1, by=c("Year", "Month"))
jobmerge<-merge(jobmerge, job_SY_aggr1, by=c("Year", "Month"))
england_AF= gtrends(keyword="england",geo="AF", gprop ="web",time= "2017-01-01 2022-12-31")
england_AF_time_trend=england_AF$interest_over_time
england_AF_time_trend$Year <- strftime(england_AF_time_trend$date, "%Y")
england_AF_time_trend$Month <- strftime(england_AF_time_trend$date, "%m")
england_AF_time_trend<-england_AF_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
england_AF_aggr1 <- aggregate(hits ~ Month + Year+keyword, england_AF_time_trend, FUN = mean)
england_AF_aggr1<-dcast(england_AF_aggr1, Year+Month~keyword)%>%rename(england_AF=england)
england_AL= gtrends(keyword="england",geo="AL", gprop ="web",time= "2017-01-01 2022-12-31")
england_AL_time_trend=england_AL$interest_over_time
england_AL_time_trend$Year <- strftime(england_AL_time_trend$date, "%Y")
england_AL_time_trend$Month <- strftime(england_AL_time_trend$date, "%m")
england_AL_time_trend<-england_AL_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
england_AL_aggr1 <- aggregate(hits ~ Month + Year+keyword, england_AL_time_trend, FUN = mean)
england_AL_aggr1<-dcast(england_AL_aggr1, Year+Month~keyword)%>%rename(england_AL=england)
england_IR= gtrends(keyword="england",geo="IR", gprop ="web",time= "2017-01-01 2022-12-31")
england_IR_time_trend=england_IR$interest_over_time
england_IR_time_trend$Year <- strftime(england_IR_time_trend$date, "%Y")
england_IR_time_trend$Month <- strftime(england_IR_time_trend$date, "%m")
england_IR_time_trend<-england_IR_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
england_IR_aggr1 <- aggregate(hits ~ Month + Year+keyword, england_IR_time_trend, FUN = mean)
england_IR_aggr1<-dcast(england_IR_aggr1, Year+Month~keyword)%>%rename(england_IR=england)
england_SY= gtrends(keyword="england",geo="SY", gprop ="web",time= "2017-01-01 2022-12-31")
england_SY_time_trend=england_SY$interest_over_time
england_SY_time_trend$Year <- strftime(england_SY_time_trend$date, "%Y")
england_SY_time_trend$Month <- strftime(england_SY_time_trend$date, "%m")
england_SY_time_trend<-england_SY_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
england_SY_aggr1 <- aggregate(hits ~ Month + Year+keyword, england_SY_time_trend, FUN = mean)
england_SY_aggr1<-dcast(england_SY_aggr1, Year+Month~keyword)%>%rename(england_SY=england)
england_IQ= gtrends(keyword="england",geo="IQ", gprop ="web",time= "2017-01-01 2022-12-31")
england_IQ_time_trend=england_IQ$interest_over_time
england_IQ_time_trend$Year <- strftime(england_IQ_time_trend$date, "%Y")
england_IQ_time_trend$Month <- strftime(england_IQ_time_trend$date, "%m")
england_IQ_time_trend<-england_IQ_time_trend%>%
mutate(hits=as.numeric(hits))%>%
replace(is.na(.),0)%>%
dplyr::select(Year, Month, keyword, hits)
england_IQ_aggr1 <- aggregate(hits ~ Month + Year+keyword, england_IQ_time_trend, FUN = mean)
england_IQ_aggr1<-dcast(england_IQ_aggr1, Year+Month~keyword)%>%rename(england_IQ=england)
englandmerge<-merge(england_AF_aggr1, england_AL_aggr1, by=c("Year", "Month"))
englandmerge<-merge(englandmerge, england_IR_aggr1, by=c("Year", "Month"))
englandmerge<-merge(englandmerge, england_IQ_aggr1, by=c("Year", "Month"))
englandmerge<-merge(englandmerge, england_SY_aggr1, by=c("Year", "Month"))
gmmerge<-merge(visamerge, jobmerge, by=c("Year", "Month"))
gmerge<-merge(gmmerge, englandmerge,by=c("Year", "Month"))
M1<-merge(FOI_Complete,Conflict_Agg, by=c("Year", "Month"))
M2<-merge(M1, Disasters_Agg, by=c("Year", "Month"))
M3<-merge(M2,Exchange_Rates_Monthly, by=c("Year", "Month"))
M4<-merge(M3,Channel_Weather_Monthly3, by=c("Year", "Month"))
M5<-merge(M4,Commodities_Dated, by=c("Year", "Month"))
M6<-merge(M5,covid_Final, by=c("Year", "Month"))
M7<-merge(M6, covid_aggr1, by=c("Year", "Month"))
M8<-merge(M7, Frontex, by=c("Year", "Month"))
M9<-merge(M8, PRCL, by=c("Year", "Month"))
M10<-merge(M9, gmerge, by=c("Year", "Month"))
M11<-merge(M10, UK_Unemployment, by=c("Year", "Month"))
M11$ID <- seq_along(M11[,1])
M11$Dublin<-ifelse(M11$ID>48, 0, 1)
DF<-M11%>%
rename(SBArrivals=`Small boat arrivals`)%>%
rename(Frontex_Detections=FrontexDetections)
write.csv(DF, "ThesisFrame1.csv", row.names=FALSE)
Data<-read.csv("ThesisFrame1.csv")
Data_Graphs<-Data%>%
rename(`Small Boat`=SBArrivals)
Dataplot<-Data_Graphs%>%
dplyr::select(`Small Boat`, Year, Month)
time.points.plot <- seq.Date(as.Date("2017-01-01"), by = "month", length.out= 72)
Dataplot1 <- zoo(Dataplot, time.points.plot)
ts_info(Dataplot1)
The Dataplot1 series is a zoo object with 3 variables and 72 observations
Frequency: monthly
Start time: 2017-01-01
End time: 2022-12-01
TSstudio::ts_plot(Dataplot1, title="Monthly Arrivals", Xtitle="Time", Ytitle="Number of Arrivals")
time.points.plot
[1] "2017-01-01" "2017-02-01" "2017-03-01" "2017-04-01" "2017-05-01"
[6] "2017-06-01" "2017-07-01" "2017-08-01" "2017-09-01" "2017-10-01"
[11] "2017-11-01" "2017-12-01" "2018-01-01" "2018-02-01" "2018-03-01"
[16] "2018-04-01" "2018-05-01" "2018-06-01" "2018-07-01" "2018-08-01"
[21] "2018-09-01" "2018-10-01" "2018-11-01" "2018-12-01" "2019-01-01"
[26] "2019-02-01" "2019-03-01" "2019-04-01" "2019-05-01" "2019-06-01"
[31] "2019-07-01" "2019-08-01" "2019-09-01" "2019-10-01" "2019-11-01"
[36] "2019-12-01" "2020-01-01" "2020-02-01" "2020-03-01" "2020-04-01"
[41] "2020-05-01" "2020-06-01" "2020-07-01" "2020-08-01" "2020-09-01"
[46] "2020-10-01" "2020-11-01" "2020-12-01" "2021-01-01" "2021-02-01"
[51] "2021-03-01" "2021-04-01" "2021-05-01" "2021-06-01" "2021-07-01"
[56] "2021-08-01" "2021-09-01" "2021-10-01" "2021-11-01" "2021-12-01"
[61] "2022-01-01" "2022-02-01" "2022-03-01" "2022-04-01" "2022-05-01"
[66] "2022-06-01" "2022-07-01" "2022-08-01" "2022-09-01" "2022-10-01"
[71] "2022-11-01" "2022-12-01"
ts<-ts(Dataplot$`Small Boat`, start=c(2017,1), end=c(2022,12), frequency=12)
frequency(ts)
[1] 12
ts_de<-decompose(ts)
plot(ts_de)
Data<-read.csv("ThesisFrame1.csv")
Data1<-Data%>%
dplyr::select(Year, Month, SBArrivals, Dublin, contains("Unemployment"), contains("Alg"), contains("Afg"), contains("Iran"), contains("Iran"), contains("Syr"), contains("Eri"), contains("Sud"), contains("Egp"), contains("Turk"), contains("Ind"), contains("Eth"), contains("Viet"), contains("Kuw"), contains("Geo"), contains("Pak"), contains("Sri"), contains("Yem"), contains("Alg"),contains("Lib"),contains("job"), contains("visa"), contains("england"), mean_temp,mean_precip,mean_wspeed, PALLFNF, PNFUEL, PNRG, PAGRI,PMETA, PFERT,C8EV_International.travel.controls, g_covid, Frontex_Detections, contains("equivalent"))%>%
dplyr::select(-one_GBP_equivalent_to_x_BHD, -one_GBP_equivalent_to_x_DJF, -one_GBP_equivalent_to_x_ILS, -one_GBP_equivalent_to_x_JOD, -one_GBP_equivalent_to_x_LBP,-one_GBP_equivalent_to_x_TND, -one_GBP_equivalent_to_x_MAD)%>%
rename(UK_Unemployment=contains("Unemployment"))
Data1$Month <- sprintf("%02d", as.numeric(Data1$Month))
DF_vars1<-Data1%>%
dplyr::select(-Year,-Month, -`SBArrivals`)
DF_split1<-Data1%>%
dplyr::select(Year,Month,SBArrivals)
lags <- seq(6)
lag_names <- paste("lag", formatC(lags, width = nchar(max(lags)), flag = "0"),
sep = "_")
lag_functions <- setNames(paste("dplyr::lag(., ", lags, ")"), lag_names)
DFLags1<-DF_vars1%>% mutate_all(funs_(lag_functions))
DF_lags1<-bind_cols(DF_split1, DFLags1)
DF_lags1 <- DF_lags1[-(1:6),]
Data4<-DF_lags1%>%
dplyr::select(-Year, -Month)
time.points.lag <- seq.Date(as.Date("2017-07-01"), by = "month", length.out= 66)
Data5 <- zoo(Data4, time.points.lag)
pre.period.lag <- as.Date(c("2017-07-01", "2022-03-01"))
post.period.lag <- as.Date(c("2022-04-01", "2022-12-01"))
impact_19 <- CausalImpact(Data5, pre.period.lag, post.period.lag, model.args=list(nseasons=12))
plot(impact_19, c("original", "pointwise", "cumulative"))+labs(x="Time", y="Small Boat Arrivals")
summary(impact_19, "report")
Analysis report {CausalImpact}
During the post-intervention period, the response variable had an average value of approx. 4.39K. In the absence of an intervention, we would have expected an average response of 1.31K. The 95% interval of this counterfactual prediction is [-0.29K, 3.40K]. Subtracting this prediction from the observed response yields an estimate of the causal effect the intervention had on the response variable. This effect is 3.09K with a 95% interval of [0.99K, 4.68K]. For a discussion of the significance of this effect, see below.
Summing up the individual data points during the post-intervention period (which can only sometimes be meaningfully interpreted), the response variable had an overall value of 39.55K. Had the intervention not taken place, we would have expected a sum of 11.75K. The 95% interval of this prediction is [-2.60K, 30.62K].
The above results are given in terms of absolute numbers. In relative terms, the response variable showed an increase of +496%. The 95% interval of this percentage is [-1473%, +4076%].
This means that, although the intervention appears to have caused a positive effect, this effect is not statistically significant when considering the entire post-intervention period as a whole. Individual days or shorter stretches within the intervention period may of course still have had a significant effect, as indicated whenever the lower limit of the impact time series (lower plot) was above zero. The apparent effect could be the result of random fluctuations that are unrelated to the intervention. This is often the case when the intervention period is very long and includes much of the time when the effect has already worn off. It can also be the case when the intervention period is too short to distinguish the signal from the noise. Finally, failing to find a significant effect can happen when there are not enough control variables or when these variables do not correlate well with the response variable during the learning period.
The probability of obtaining this effect by chance is very small (Bayesian one-sided tail-area probability p = 0.005). This means the causal effect can be considered statistically significant.
summary(impact_19)
Posterior inference {CausalImpact}
Average Cumulative
Actual 4394 39545
Prediction (s.d.) 1306 (1609) 11753 (14485)
95% CI [-289, 3402] [-2602, 30622]
Absolute effect (s.d.) 3088 (1609) 27792 (14485)
95% CI [991, 4683] [8923, 42147]
Relative effect (s.d.) 496% (4607%) 496% (4607%)
95% CI [-1473%, 4076%] [-1473%, 4076%]
Posterior tail-area probability p: 0.00502
Posterior prob. of a causal effect: 99.49799%
For more details, type: summary(impact, "report")
plot(impact_19$model$bsts.model, "coefficients",inclusion.threshold=.01 ,cex.names=.4)
